Hugging Face
Models
Datasets
Spaces
Community
Docs
Enterprise
Pricing
Log In
Sign Up
Spaces:
lym0302
/
DeepSound-V1
like
0
Running
App
Files
Files
Community
Fetching metadata from the HF Docker repository...
main
DeepSound-V1
/
third_party
/
VideoLLaMA2
/
videollama2
/
model
Ctrl+K
Ctrl+K
2 contributors
History:
12 commits
lym0302
frames_features = frames_features.to(dtype=torch.float16)
4c6ac06
about 2 months ago
beats
our
2 months ago
__init__.py
Safe
11.1 kB
bfloat16
2 months ago
encoder.py
7.19 kB
our
2 months ago
projector.py
Safe
9.72 kB
x = x.to(torch.float16) # bfloat16
2 months ago
videollama2_arch.py
Safe
18.5 kB
frames_features = frames_features.to(dtype=torch.float16)
about 2 months ago
videollama2_gemma2.py
Safe
6.25 kB
our
2 months ago
videollama2_llama.py
Safe
5.46 kB
our
2 months ago
videollama2_mistral.py
Safe
5.55 kB
our
2 months ago
videollama2_mixtral.py
Safe
5.38 kB
our
2 months ago
videollama2_phi3.py
Safe
5.49 kB
our
2 months ago
videollama2_qwen2.py
Safe
5.4 kB
our
2 months ago